In almost every imaginable scenario, a clean energy future will include large-scale solar panel installations. Yet manufacturing the silicon metal required for them is complex and resource intensive.
Northwest communities have been fighting an onslaught of dirty energy proposals for nearly a decade, from coal terminals and oil pipelines to petrochemical refineries and natural gas facilities. Many of these projects marketed themselves as environmentally responsible, but they were all, in one way or another, expansions of the fossil fuel industry. They were dirty.
But what to make of a big industrial project proposal—one that uses coal and creates pollution—that would manufacture a key component of the region’s clean energy future?
That’s the question on the table in northeast Washington’s Pend Oreille County, where a Canadian mining company is looking to construct a $325 million smelter. The facility would use a small amount of coal, about 50,000 metric tons each year, to melt down sand to pure silicon for solar panels and other electronics. Though the project’s development is several months behind its original schedule, the proposal has kicked off a debate about the tradeoffs involved in the Northwest’s transition to renewable energy.
Opponents include residents in a rural area along the Idaho-Washington border, as well as a nearby tribe worried about air pollution and traffic congestion. But the project’s backers argue that because much of that silicon will be used to produce solar panels, it will actually help to improve air quality and fight climate change.
A type of semiconductor, pure silicon metal is a key component not only in solar panels, but also in computer chips and other products with electronic circuits. The company has yet to confirm any official customer contracts, but officials point out that it could deliver silicon metal to REC Silicon in Moses Lake, which would further refine the metal into high-purity polysilicon for its made-in-Washington solar panels.
In almost every imaginable scenario, a clean energy future will include large-scale solar panel installations. Yet manufacturing the silicon metal required for them is complex and resource intensive. The proposed project in northeast Washington is no exception.
What it takes to make a solar panel
Turning silica (aka silicon dioxide) into pure silicon metal isn’t easy. To do it, electric furnaces at the smelter must generate temperatures around 3,000 degrees Fahrenheit to melt a silica-fuel mixture and catalyze a chemical reaction to isolate the silicon metal. Even at their most efficient, these furnaces would have a voracious appetite for electricity: around 105 megawatts on a continuous basis, roughly the equivalent of 68,000 homes. Although the smelter would ostensibly be powered by Pend Oreille Public Utility District’s Box Canyon Dam, the facility would demand more power than the dam could provide on its own. (It has a maximum nameplate capacity of 90 megawatts after a 2015 turbine upgrade.) In fact, it’s more than four times the amount of electricity used by all of the utility’s residential and small commercial customers combined.
Producing one ton of silicon metal requires about six tons of raw materials, so HiTest’s planned 73,000 metric tons of silicon per year would require the delivery of hundreds of thousands of tons of raw materials to the Newport site. Nearby saw mills would send seven or eight trucks per day to deliver wood chips, which are integral to the smelting process. Trains loaded with silica from the Canadian quarry HiTest owns would share the tracks with others delivering coal to be mixed with the silica in the furnaces. The smelting process requires a rare type of metallurgic coal (likely obtained from Kentucky) known as “blue gem,” capable of burning at high temperatures with low ash and sulfur content so as not to contaminate the silicon metal. Operations at the smelter would demand approximately 48,000 metric tons of coal per year—roughly 40 rail cars each month.
All that assumes that the company can convince BNSF Railway to extend a railroad spur to the smelter site. If not, HiTest might instead use an estimated 37 trucks per day to transport the raw materials the last few miles from Idaho. Some residents contend that the smelter’s operations would require many more, but the company disputes these higher figures.
Yet the project’s impact on the public extends beyond local traffic concerns. The plan also faces questions about whether taxpayers are subsidizing the project.
The smelter and its discontents
In 2016, the Washington Department of Commerce designated the smelter proposal a “Project of Statewide Significance” and awarded HiTest a $300,000 economic development assistance grant. In September 2017, Pend Oreille Public Utility District (PUD) sold 186 acres in four parcels near Newport to the company for the same amount—and opponents of the project cried foul.
A group of local residents formed Citizens Against the Newport Smelter (CANSS), which has filed a lawsuit challenging the land deal between Pend Oreille PUD and HiTest. CANSS claims that the utility illegally bought 14 acres of land from the county with the intention of bundling it with other parcels for sale to HiTest just a week later. Further, the group alleges that the utility failed to follow procedure by declaring the land “surplus” (which would have triggered a public notification process and a solicitation of bids), rather than negotiating exclusively with HiTest for a sale. Pend Oreille PUD disputes that characterization, asserting that it posted the land for sale in the local paper multiple times, in compliance with state law.
CANSS also points out that the HiTest land is currently zoned as public land, meaning it would have to be rezoned before it could be used for private industrial purposes, which would in turn require action by the county commissioners and county planning commission, along with public hearings.
HiTest will also benefit from cheap electricity provided by the Box Canyon Dam, about 55 miles north of Newport. Exactly how much HiTest will pay is not clear because the company and the utility must still negotiate a power supply agreement. But at least one Canadian official, former BC Mines Minister Bill Bennett, has accused the county of offering below-cost electricity in order to woo the project south of the border.
Yet Pend Oreille County PUD’s rate schedule stipulates that contracts for large industrial customers must recover “all increased power costs arising from the new contractual load”—meaning that these enterprises would be required to pay their own way—and HiTest is currently funding a study of the potential impacts on the electrical grid. For its part, the PUD may be eager to find a big new industrial customer given uncertainty surrounding an old one: Ponderay Newsprint, which now consumes a remarkable 74 percent of the PUD’s power load, is forecasting reduced electricity demand and may even cease operating entirely in the near future owing to falling demand for newsprint.
The nearby Kalispel Tribe, whose traditional homeland spans present day northeastern Washington, northern Idaho, and northwestern Montana, opposes the project, arguing that the smelter’s air pollution and climate impacts are too severe and pointing out that the tribe’s most important natural resources have been historically degraded by government policy.
Meanwhile, Spokane-based environmental advocacy group The Lands Council objects to the modeling methods HiTest relied on for its air pollution estimates, which were not specific to the unique topography and weather patterns of the Pend Oreille River Valley. Neighboring tribes, including the Spokane and Kootenai, have registered similar concerns, calling for better air quality modeling techniques, data-sharing, and greater attention to native sovereignty before the state issues permits. In January 2018, the Affiliated Tribes of Northwest Indians, the major regional consortium of tribes, published a joint statement calling for a halt to the permitting process and greater scrutiny of the project. And in April 2018, the Pend Oreille County Commissioners also called for more locally-specific air quality modeling and a formal health impact study.
The citizens group, CANSS, alleges that air pollution from the smelter could create potential health hazards. The group points to high levels of arsenic near a silicon smelter in Iceland as a cautionary tale.
We are a nonprofit. Donate now to support more research like this!
The company’s own estimates show that silicon metal production at the site would indeed produce pollution—a little more than 320,000 metric tons of carbon dioxide per year, making it the 15th largest emitter in the state, ahead of the Wenatchee Alcoa aluminum smelter and the US Oil refinery in Tacoma. In addition, HiTest predicts it will generate over 700 metric tons per year each of smog- and acid rain-causing sulfur dioxide and nitrogen dioxide. It would be the 6th largest emitter of sulfur dioxide in the state, and the 13th largest emitter of nitrogen dioxide. On the other hand, air quality in the Pend Oreille Valley is quite good, so while new pollution sources might mark a change from the status quo, overall pollution levels would likely remain relatively low.
All those figures would increase if the smelter doubles its furnace capacity, something HiTest is considering if market conditions prove favorable.
HiTest faces several hurdles before it can break ground on the project. Based on the company’s original timetable, the project is at least nine months behind schedule and the project backers have yet to file a permit application with the Washington Department of Ecology, the agency that would oversee review and permitting. Once they do, the enterprise will be subject to an environmental review under the State Environmental Policy Act, an assessment that will inform decisions about key air quality permits. Plus, the smelter will also need a building permit and a zoning permit from Pend Oreille County.
Clean energy questions
Whatever happens, the silicon smelter proposal embodies the tradeoffs of fighting climate change. Do nothing, and manmade ravages to the planet’s atmosphere will condemn millions of people to displacement and economic impoverishment. Yet many of the solutions to climate change sometimes inflict their own damage. Much like carbon taxes can unduly burden low income people, so too industrial development for clean energy purposes may have impacts on the most powerless, whether largely white rural communities or historically oppressed indigenous peoples like the Kalispel Tribe.
The Northwest is rapidly transitioning to clean energy sources, including solar panels made with polysilicon that essentially require mining and smelting to happen somewhere on earth. Air pollution from a silicon smelter may affect nearby communities, but blocking a new smelter in Cascadia will likely only serve to drive solar manufacturing to other countries, perhaps in places with fewer environmental safeguards.
Can Northwest leaders figure out a way to mitigate the project’s pollution, so that they might permit it without undue impact to nearby communities and tribes? And if they can’t, what then for the future of silicon metal manufacturing and, by extension, solar panel-generated electricity? Unlike so many of the coal, oil, and gas projects contested by the Thin Green Line, the region’s opposition movement to fossil fuels, the Newport silicon smelter offers no easy answers.
Throughout this article we refer to the project backers as HiTest Sand, consistent with usage up until the company’s recent rebranding of its smelter-development arm as PacWest Silicon.
Addendum 6/26/18: Between the time when we researched & wrote this article and when it published, the projected backer submitted a formal letter to the Washington Department of Ecology. Based on information in that letter, we amended this article to reflect the larger 73,000 ton production volume (which we initially reported as 60,000 consistent with what was available in the public record at the time).
Also of interest in the PacWest letter:
- 50 percent of the silicon metal produced at the facility is earmarked for solar cells.
- HiTest gives a little more detail about their raw material inputs: 170,000 tons/year quartz (silica), 150,000 tons/year coal + charcoal, and 130,000 tons/year wood chips. It’s not clear what the precise breakdown between the coal and charcoal would be, so we can’t say whether it would be any different from the 48,000 tons/year coal we estimated. (However, the numbers in the submission to Ecology don’t quite square with the ratios mentioned in the company’s PowerPoint—in particular, the coal + charcoal number seems high relative to the wood chip number.) Altogether, and assuming the 6:1 ratio of raw materials the company mentioned in the PowerPoint presentation, the volume of inputs imply a production level of roughly 75,000 tons/year of silicon metal production—not far from 73,000 the firm mentions in the letter to Ecology. That would be equal to the current largest production facility in the US, which is Ferroglobe Alloy in West Virgina.
- The company estimates approximately 50 commercial/non-passenger vehicle trips to the facility each day. That doesn’t mean trucks per se, and the revised figure could still be consistent with their 37 truck per day estimate.There’s just not enough information to go on. The letter does, however, seem to indicate that the vehicle figures are independent of whether or not the rail spur is constructed.
- The company estimates an average of 10 rail cars per day from all raw materials combined. This is not necessarily inconsistent with our 40/month estimate for coal alone.
Ahren Stroming is a research contributor to Sightline. He’s a former policy analyst for the City of Seattle Office of Sustainability and Environment and a current graduate student at the University of Freiburg in Germany.